Therapeutic drug monitoring
Most drugs are dosed by a simple approximation — a standard amount for a 70kg adult, adjusted a bit for age. For most drugs, in most patients, the therapeutic window is wide enough that this approximation is safe. For a handful of drugs, it isn't.
Take tacrolimus, given to transplant patients to stop organ rejection. Too little, and the immune system attacks the new organ. Too much, and the drug itself damages the kidneys and raises infection and cancer risk. The gap between "too little" and "too much" is a concentration range narrow enough to fit on one graph, and the dose needed to land in that range can vary ten- to twentyfold between individual patients. You can't guess this dose safely — you have to measure it.
Therapeutic drug monitoring (TDM) is the practice of measuring a drug's concentration in a patient's blood, usually at a specific time relative to the last dose, and using that measurement to adjust their individual dose toward a target range associated with good outcomes. The classic TDM drugs are tacrolimus and cyclosporine (transplant), vancomycin and the aminoglycosides (serious infection), phenytoin and carbamazepine (epilepsy), lithium (bipolar disorder), methotrexate, and digoxin (heart failure).
The modern version of this — model-informed precision dosing (MIPD) — uses a Bayesian algorithm that combines a patient's own measured levels with a population PK model (see pharmacometrics) to predict that specific patient's PK parameters and calculate an individualised dose in close to real time.
Where it sits in development
Unlike most of the rest of this pillar, TDM sits almost entirely at the post-approval, clinical-use end of the pipeline — it's practised in hospitals, transplant centres, and infectious disease units, not mainly in a drug-development lab. Earlier-stage work still feeds it: Phase I/II data identifies whether a drug needs TDM at all (is the therapeutic window narrow? is patient variability high?), and the target concentration range that TDM is measured against usually comes out of the same exposure-response analysis that sets the drug's dosing recommendation.
What's actually settled, and what isn't
Classical trough-level TDM for drugs like digoxin and phenytoin has been routine clinical practice for more than fifty years. What's moving now:
- MIPD at scale. The science of Bayesian dose individualisation is proven — the constraint is adoption. Most hospitals still don't use it even for vancomycin, where the clinical case for switching from trough-only monitoring to full exposure-guided dosing is well established.
- TDM for biologics — monitoring blood levels of drugs like infliximab or adalimumab in inflammatory bowel disease — has real evidence behind it for a few specific drugs, but target ranges are still poorly defined for most biologics, and clinical uptake is inconsistent.
- TDM in critically ill patients is a distinct problem: fluid shifts, altered protein binding, and augmented renal clearance mean a model built on stable outpatients doesn't transfer cleanly to the ICU.
- Point-of-care measurement — getting a drug level back in minutes at the bedside instead of hours from a central lab — would make TDM practical for drugs where it currently isn't, but the technology isn't there yet for most of the classic TDM drugs.
- Pediatric TDM target ranges are, more often than not, extrapolated from adult data rather than established directly in children.